ultrasonic signal processing toolbox user manual v1 · the ultrasonic signal processing toolbox is...

ULTRASONIC SIGNAL PROCESSING TOOLBOX

User Manual v1.0

Acknowledgment

The authors would like to acknowledge the financial support of European Commission within the project SPIQNAR FIKS-CT-2000-00065

copyright

Lars Ericsson

Uppsala University

Signals and Systems

September 2001

Disclaimer The Ultrasonic Signal Processing Toolbox (USPT) can be used for research purposes only. It is distributed in the hope that it will be useful for the research community but WITHOUT ANY WARRANTY.

More specifically:

THE PROGRAM IS PROVIDED “AS-IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE IMPLIED WARRANTIES OR CONDITIONS OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT SHALL ANY OF THE AUTHORS OF THE USPT AND/OR THE DEPARTMENT OF ENGINEERING SCIENCES AT UPPSALA UNIVERSITY, SWEDEN, BE LIABLE FOR ANY SPECIAL, INCIDENTAL, INDIRECT, OR CONSEQUENTIAL DAMAGES OF ANY KIND, OR DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA, OR PROFITS, WHETHER OR NOT THE AUTHORS OF THE DREAM TOOLBOX HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES, AND/OR ON ANY THEORY OF LIABILITY ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

RESTRICTED

September 2001 iii SPIQNAR

CONTENTS

1. INTRODUCTION 1

2. INSTALLATION 1

2.1 SYSTEM REQUIREMENTS 12.2 INSTALLING THE SOFTWARE 1

3. GETTING STARTED 1

4. FILE FORMATS 2

4.1 SUPPORTEDFORMATS 24.2 ADDING NEW FORMATS 3

5. SIGNAL PROCESSING FUNCTIONS 3

5.1 NONCOHERENTDETECTION 35.2 COMMON COMPONENTREJECTION 65.3 SPLIT SPECTRUMPROCESSING 7

5.3.1 Consecutive Polarity Coincidence 75.3.2 Amplitude Minimization 9

6. RECOMMENDATIONS FOR ALGORITHM USE 9

RESTRICTED

September 2001 SPIQNAR1

1. INTRODUCTION

This manual presents the first end user release of the Ultrasonic Signal Processing Toolboxfor early evaluation of the material noise suppression algorithms proposed for the SPIQNARproject. The software is operated from a graphical user-interface and has been designed to bemore or less self-documenting and easy to use for NDT operators.

Currently four signal processing algorithms are included: Noncoherent Detection, CommonComponent Rejection, Split Spectrum/Consecutive Polarity Coincidence and SplitSpectrum/Minimization. More algorithms may be included later in the project.

The software supports a number of file formats for reading, of which a simple MATLABformat is described in detail below. New formats may be added by modifying functionssupplied as source code. Alternatively a stand-alone software that converts to any of thesupported formats may be used.

2. INSTALLATION

2.1 System Requirements

The Ultrasonic Signal Processing Toolbox (USPT) requires a Pentium compatible PC runningMicrosoft Windows 2000, NT 4.0, 98 or ME and MATLAB 5.3 or later. The computershould be equipped with at least 128 MB of RAM for the software to run smoothly.

It is assumed throughout this manual that MATLAB is properly installed and that the user isfamiliar with using Microsoft Windows.

2.2 Installing the Software

The software is supplied in a zip-archive containing the required MATLAB files. In order toinstall the software create a new directory on the hard disk and unpack these files into thedirectory. The new directory must then be added to the MATLAB path. This is most easilyperformed by clicking on the Path Browser button in the MATLAB command window.

3. GETTING STARTED

The Ultrasonic Signal Processing Toolbox is started by first starting MATLAB and thentyping ultra on the command line. The graphical user interface shown in Fig. 1 will thenappear. By using the graphical controls the operator can load and view US data, performsignal processing and save the result. Signal processing parameters may also be saved forlater use on similar data.

The software handles RF B-scan data only so 3D data volumes have to be loaded, viewed andprocessed B-scan by B-scan. After loading a B-scan it can be viewed using different colourmaps and amplitude scaling. The presentation can also be switched between RF, rectified and

RESTRICTED


envelope. The sliders at he bottom and right of the B-scan are used for positioning cursors andselecting what A-scan to view.

Fig. 1. Graphical user interface of the Ultrasonic Signal Processing Toolbox.

4. FILE FORMATS

4.1 Supported Formats

As distributed the USPT software can read ultrasonic data from three file formats:

ÿ� Native MATLAB (.mat)ÿ� Ultra Optec (.scn)ÿ� NDT Systems (.ndt)

The NDT Systems-format supports 3D data volumes while the other support B-scans only.When using the MATLAB format the mat-file has to include a variable namedusDatacontaining a B-scan with A-scans as columns, and a variable namedsampFreq containingthe sampling frequency in MHz. This is also the only format that can be used for saving theresults of processing.

RESTRICTED


4.2 Adding New Formats

Adding new file formats for reading can easily be done by the user as the file loadingfunctions are supplied as MATLAB source files (m-files). There are two files that need to beconsidered:

ÿ� loaddata.m : Loads the first B-scan from a file. If the format supports a single B-scan, only this function needs to be modified.

ÿ� loaddatb.m : Loads a B-scan with a specified number from a 3D data volume.

Instructions for how to add code for actually reading the US data can be found in these files inthe sections starting withelseif strcmpi(extension,’xxx’) , wherexxx shouldbe replaced by the desired file name extension.

5. SIGNAL PROCESSING FUNCTIONS

Four signal processing algorithms have been included in USPT v1.0. The most well knownprocessing scheme for ultrasonic grain noise suppression, Split Spectrum Processing (SSP), isrepresented by the traditional Minimization algorithm and a modified Polarity Thresholdingalgorithm referred to as Consecutive Polarity Coincidence. These algorithms are included forcomparison but can not be recommended for regular use due to the inherent difficulties withproper calibration.

The recommended algorithm to use when a relevant reference block is available forcalibration is the Noncoherent Detection. An automatic tuning procedure has been included tomake calibration easy. The fourth algorithm, Common Component Rejection, could be worthtrying if anomalies, i.e., defects, are to be found in objects when calibration is not feasible.Virtually no parameters are needed in this case.

5.1 Noncoherent Detection

Thenoncoherent detector(NCD), which is well known from communications theory, isdesigned for detection of bandpass signals,s(t)=A(t)cos(2πf0+φ), in additive Gaussian noise.As the term noncoherent implies, the algorithm is capable of detecting signals with unknownphase,φ. From a NDT perspective, the noncoherent detector is designed to detect a family oftransients defined by the set of transients obtained by continuously varying the angleφ overthe interval [0,2π). Since it is impossible to exactly specify what transients to expect afterpropagation and reflection in NDT, the design of family detectors seems attractive.

Application of the NCD algorithm in ultrasonics differ from telecommunication in that notonly the phase,φ, is unknown but also the envelope,A(t), and centre frequency,f0, of thetransients. Thus, the ultrasonic response signals must be modelled by a transient prototypethat can be adapted to the inspection set-up, and which is general enough to be useful for alltarget echoes within the inspected volume. In the USPT software a Gaussian shaped transientmodel with an adjustable centre frequency and bandwidth has been used.

RESTRICTED


The NCD algorithm includes specification of the transient prototype and estimation of theautocorrelation function of the grain noise, followed by the computation of an optimal filter.The USPT software implements an automatic procedure for finding proper values for theseprototype parameters, consisting of the following steps:

ÿ� Load a B-scan from a reference block with known defect locations. The defect usedfor calibration should preferably be “soft”, i.e. a side-drilled hole, located at a similardepth as the volume to be inspected in the target object.

ÿ� Click on Mark in the Noise Region Selection area and then use the mouse to “draw” arectangle around a region in the B-scan containing noise only. The noise region shouldbe around the same depth as the reference defect and large enough to provide goodstatistics. The coordinates of the selected region will be displayed to the left of thebutton.

ÿ� Click on Mark in the Target Selection area and use the mouse in the same way asabove to select the reference defect.

ÿ� In order to achieve acceptable computation time during the search for proper prototypeparameters the frequency space should be restricted to a relevant region, bounded bythe lowest and highest frequencies in the Filter Parameters area, together with theresolution given by the Frequency Step. The Minimum Bandwidth should be wideenough to give a time resolution required by the inspection, but at the same timenarrow enough to allow the calibration procedure to find the best parameters.Examples of filter parameters for a wide-band 1.8 MHz transducer, used for inspectionof cast stainless steel, are shown in Fig. 2.

ÿ� The filter calibration procedure is started by clicking onCalc. Filter . A windowis then opened showing the noise autocorrelation function, each transient prototypethat is evaluated, the corresponding NCD filter impulse response and the enhancementof the signal-to-noise ratio (SNR) for each filter. The SNR calculation is performed bydividing the signal energy in the selected target region by the energy in the noiseregion. The transient parameters corresponding to the highest SNR enhancement willbe selected after the process is completed.

ÿ� After the parameter search process has been finished the NCD window has to bemanually closed. The best filter obtained can then be employed to the reference B-scan by clicking onProcess . The last step in the NCD calibration procedure is to setthe filter gain using the slider below the image. The filter design is now completed andthe filter can be saved for future use on US data from similar materials with defectzones on approximately the same depth, just by clicking onProcess . An example ofthe result after processing is shown in Fig. 3.

RESTRICTED


Fig.2 Design of Noncoherent Detection filter.

Fig. 3 Result after processing by Noncoherent Detection.

RESTRICTED


5.2 Common Component Rejection

The common component rejection is a very simple algorithm, based on a noise model only,that may be useful when no reference data is available for calibration. It is based on theassumption that the grain noise is spectrally different from defect responses, more or lessstatistically stationary and far more common than defect indications. By suppressing thosefrequency components that dominate the B-scan and enhance non-common behaviour thenoise will be suppressed and hopefully the defects will remain. This algorithm is onlyintended for interactive blind search for defects and not for regular inspection use.

In order to calculate a filter a noise region has to be marked in the same way as forNoncoherent Detection. Then, the only parameter that has to be supplied before clicking onCalc. Filter is the filter length inµs. It should be at least as long as the transducerimpulse response but be kept as short as possible to avoid a large “dead zone” in theprocessed B-scan. The length is appropriate when the amplitude of the resulting filterresponse comes close to zero at the end-points, which can be seen in the window that isopened during the filter calculation, see Fig. 4 for an example. The Low-pass Filteringcheckbox should normally be checked to remove the high frequency receiver and quantizationnoise that is otherwise amplified by the filtering process. If there is any reason to believe thatdefect information is present at frequencies higher than in the grain noise the box should beunchecked. The resulting filter is used in the same way as for Noncoherent Detection.

Fig.4 Design of Common Component Rejection filter.

RESTRICTED


5.3 Split Spectrum Processing

TheSplit Spectrum Processing(SSP) algorithm consists of a bank of Gaussian bandpassfilters for signal expansion and a statistical processor for target echo extraction. The filterbank is used for obtaining a set of signals, a kind of time-frequency representation, amongwhich the noise components have been more or less decorrelated, while at the same timeretaining a large degree of correlation for target echoes. Target echo extraction may then beimplemented by applying a suitable measure of correlation to the generated set of signals.

Simple target extractors such as theAmplitude Minimizationand thePolarity Thresholdingalgorithms have been suggested and proven successful if the processing parameters have beencorrectly tuned. Both algorithms rely on the assumption that target echo information is presentin all frequency bands. Consequently, the far most important parameters, using these targetextractors, are the lower and upper cut-off frequencies of the processing range. Even a slightdeviation in the frequency range may result in target echo cancellation, since an output signalfrom a single filter containing no target information is sufficient for regarding the echo asmaterial noise. Therefore, using these target extractors requires good a priori knowledge ofthe target echo spectrum for achieving adequate results. Unfortunately, the required data is noteasily obtainable, as it depends, among other factors, on the defect properties. Consequently,calibrating the parameters against a known reference may not be feasible.

The USPT software includes two SSP-algorithms: Amplitude Minimization and an extendedversion of Polarity Thresholding, referred to asConsecutive Polarity Coincidence.

5.3.1 Consecutive Polarity Coincidence

Consecutive Polarity Coincidence (CPC) may be viewed as an adaptive version of thePolarity Thresholding algorithm. Polarity Thresholding (PT) is based on the assumption thatthe ultrasonic target echo is spectrally similar to a mathematical pulse, i.e., for a specifiedfrequency range the echo is composed of in-phase Fourier components. Prior to application ofthe PT, the received ultrasonic signal is therefore decomposed by the Split Spectrum filterbank, which generates a time-frequency representation of the signal. The conventional PTdetection algorithm utilizes the pulse characteristics by indicating a target echo at thoseinstants of time where all the split signals share a common polarity. Obviously, the underlyingassumption can be very easily violated, by adding a single split signal with the oppositepolarity.

In a sense, the PT algorithm measures the resemblance to a predetermined hypothetical pulse,at each instant of time. However, the polarity of the split signals can be used to measure the"pulseness" directly, without assuming a certain frequency range for the hypothetical pulse.For a certain instant of time, the polarities of the split signals originating from subsequentfrequency bands constitute a binary sequence. By searching for the maximum lengthsubsequence of consecutive polarity coincidences, one can obtain a direct measure of thespectral location of the best pulse that can be constructed within the processed frequencyrange. The length of the subsequence will correspond to the bandwidth of the pulse, and thestarting index to the lower cut off frequency. For example, if the filter bank consists of 30filters and the target echo is present at a certain instant of time in the split signals No. 5-27,then the conventional PT algorithm will most probably fail, while the CPC will report thepresence of a pulse, i.e., a presumed target, with a bandwidth corresponding to 22 times the

RESTRICTED


filter frequency distance. Hence, the CPC algorithm decreases the need for careful tuning ofthe frequency range used for signal expansion. The output from the CPC process whichindicates the "instantaneous bandwidth" can then be compared to a threshold level in order tocreate a gating signal that indicate presumed presences of a flaw signals.

The parameters that have to given to apply CPC are the spectral location of the filter bank(lowest and highest frequency), the number of filters and the filter bandwidth. Examples ofparameter settings for the same transducer-material combination as used with NoncoherentDetection are shown in Fig. 5. After the filtering computations have been performed byclicking onProcess the coherence level can be interactively set to an appropriate value. Ifthe coherence threshold is set to 1 CPC becomes identical to the PT algorithm. It should benoted that the CPC algorithm requires quite a narrow filter bandwidth, which implies lowtemporal resolution, i.e. closely separated targets can not be detected.

Fig.5 Result after processing by Split Spectrum / Consecutive Polarity Coincidence.

RESTRICTED


5.3.2 Amplitude Minimization

The amplitude minimization algorithm measures the correlation between the signals generatedby the SSP filter bank by finding the minimum amplitude for each instance of time amongthem. If a target is present the amplitude should be high in all frequency bands, while grainnoise only should result in random amplitudes, some of them probably close to zero.Consequently, the result will be that grain noise is suppressed compared to target echoes. Theparameters are the same as for the CPC algorithm except that the coherence threshold isreplaced by a filter gain, see Fig. 6 for an example. The low-pass filtering option shouldnormally be enabled to remove high frequency noise introduced by the minimizationprocedure.

Fig.6 Result after processing by Split Spectrum / Consecutive Polarity Coincidence.

6. RECOMMENDATIONS FOR ALGORITHM USE

The algorithms presented above may all produce very good results in specific applications.However, that is not the only criterion that is important. If the algorithms are to be regularlyused in an industrial environment it is also at least as important that they can be properlycalibrated, that they behave reasonable even if defect responses and noise characteristicsdeviate slightly from what is assumed, and that they are well theoretically understood. Theonly algorithm, among those included in USPT, that can be said to adhere to theserequirements is the Noncoherent Detection. Therefore, this algorithm is recommended fornormal inspection situations.

RESTRICTED


Noncoherent Detection is based on explicit modelling of both noise and expected flawresponses, i.e., a reference block with known defects is required for calibration. Although thisshould normally be the case there may be situations where no reference is available. Fordealing with such cases the algorithm referred to as Common Component Rejection has beenincluded in the USPT. The filter designed using this method is based on (approximate)modelling of noise only. The results produced by this filter can be useful for detectinganomalies, which can be assumed to be defects, but the peak amplitudes bear littlesignificance.

The Split Spectrum algorithms can produce very spectacular results, virtually removing allgrain noise in some cases. It should be noted though that these results are achieved byemploying highly nonlinear operations that lead to inherent difficulties with calibration,robustness, time resolution, and theoretical analysis. Although the SSP algorithms can not berecommended for practical use within this project the fact that they are well established in theNDT signal processing community motivates the inclusion in the USPT software. It may alsobe interesting to compare the performance of SSP and NCD.

ultrasonic signal processing toolbox user manual v1 · the ultrasonic signal processing toolbox is...

Documents